Remotely sensed data have distinct benefits for ecological monitoring and modeling with multiple sensors covering long time spans. The products derived from remotely sensed imagery provide information about the surface of the Earth that may not be present in other environmental datasets. Ecological niche modeling is in a unique position to utilize these data while modeling the fundamental niche of a species. Despite the benefits, use of remotely sensed data in ecological niche modeling has been rare. Native pollinating insects are vital to both ecosystem health and the agricultural industry. Knowing the potential distribution of these insects via modeling of their fundamental niche will allow future land-use planning to conserve optimum habitat. Using biophysical datasets generated from Landsat and MODIS imagery products, species distributions were generated for two ecologically important native pollinators (Bombus pensylvanicus and Bombus terricola). MaxENT was the predictive distribution modeling program used; it relies on a maximum entropy technique and has shown success in past studies. Jackknife operations were used to assess the contribution of remotely sensed data layers to prediction of species distributions and compare the results with those obtained with traditional environmental data layers.
Results/Conclusions
Preliminary results indicate that incorporating remotely sensed datasets improves output of the predictive models, but with a few considerations. Some datasets had strong associations with sampling biases (e.g. tendency to sample near roads) and resulted in over-predicted distributions. These datasets tended to be strongly significant in initial models for both study species and were removed in order to allow climate, environmental and ecosystem datasets to drive model outcomes. Unbiased datasets that best improved model results varied between species. This poster will present the final model outputs and discuss the expanded role that remotely sensed data may have in ecological niche modeling in the future.
